PERSPECTIVE

Regulation of the immune system bybiodiversity from the natural environment:An ecosystem service essential to healthGraham A. Rook1

Centre for Clinical Microbiology, Department of Infection, and the National Institute for Health Research and University College LondonHospitals Biomedical Research Centre, University College London, London NW3 2PF, United Kingdom

Edited by Ruslan Medzhitov, Yale University School of Medicine, New Haven, CT, and approved October 1, 2013 (received for review July 23, 2013)

Epidemiological studies suggest that living close to the natural environment is associated with long-term health benefits including reduceddeath rates, reduced cardiovascular disease, and reduced psychiatric problems. This is often attributed to psychological mechanisms, boostedby exercise, social interactions, and sunlight. Compared with urban environments, exposure to green spaces does indeed trigger rapidpsychological, physiological, and endocrinological effects. However, there is little evidence that these rapid transient effects cause long-termhealth benefits or even that they are a specific property of natural environments. Meanwhile, the illnesses that are increasing in high-incomecountries are associated with failing immunoregulation and poorly regulated inflammatory responses, manifested as chronically raisedC-reactive protein and proinflammatory cytokines. This failure of immunoregulation is partly attributable to a lack of exposure to organisms(“Old Friends”) from mankind’s evolutionary past that needed to be tolerated and therefore evolved roles in driving immunoregulatorymechanisms. Some Old Friends (such as helminths and infections picked up at birth that established carrier states) are almost eliminatedfrom the urban environment. This increases our dependence on Old Friends derived from our mothers, other people, animals, and theenvironment. It is suggested that the requirement for microbial input from the environment to drive immunoregulation is a major componentof the beneficial effect of green space, and a neglected ecosystem service that is essential for our well-being. This insight will allow greenspaces to be designed to optimize health benefits and will provide impetus from health systems for the preservation of ecosystem biodiversity.

Numerous studies demonstrate that livingclose to the natural rural or coastal environ-ment, often denoted “green space or “bluespace,” respectively, is beneficial for humanhealth. It reduces overall mortality, cardiovas-cular disease, and depressive symptoms andincreases subjective feelings of well-being (1–8). The beneficial effects are particularlyprominent in individuals of low socioeco-nomic status (1–3, 8). It is often suggestedthat the mechanism of this effect is psycho-logical. Looking at green spaces or walkingin parkland or forests cause rapid psycho-logical and physiological changes that canbe demonstrated not only by psychologicaltesting (4) but also by mobile elecroence-phalograms (5) and by measurements ofcerebral blood flow, various cardiac param-eters, blood pressure, and salivary cortisol(6, 7). Even looking at the natural environ-ment as images or through a window is saidto have beneficial effects (9). Some authorsexplain this from an evolutionary perspective(10). The relaxation and satisfaction derivedfrom the natural environment might repre-sent the equivalent of “habitat selection” inother species (11). As shown recently by an-alyzing carbon isotopes in tooth enamel, fromas early as 3–4 Mya, hominins were evolvingin wooded grassland (12) and followed riversand coastlines or settled near lakes. Thus,humans will have evolved to obtain psycho-logical rewards from approaching these idealhunter-gatherer habitats (10).

This psychological explanation is oftensupplemented by other factors: social inter-actions, exercise, and sunlight. For example,the natural environment might promote socialinteractions and a sense of community (13)when the natural environment is an impor-tant contributor to the social capital of theindividual (14). Similarly green spaces some-times encourage physical activity (15), al-though city-dwellers can walk to mostresources and tend to do so, whereas individ-uals living in leafy suburbs are often forced touse their cars to get anywhere at all, so exer-cise can paradoxically decrease (16, 17).Sunlight is thought to counteract seasonalaffective disorder (SAD) and has been used totreat tuberculosis and heal infected wounds(18, 19).Although all these factors may contribute

to the beneficial effects, there are two majoruncertainties about the psychological com-ponent. First, there is the issue of specificity.Most psychological studies fail to includeappropriate controls. It is not sufficient tocompare exposure to a city street withexposure to a green space. Would any suitablyrelaxing environment—a quiet comfortablecafé, or a cinema showing a feel-good film inthe urban environment—have the same psy-chological effects as green space when testedin comparison with a busy city street? [Theother suggested benefits—social interaction,exercise, and sunlight—are clearly not spe-cific for green space. Social capital usually

derives from urban social interactions, andit has not been possible, using currentlyavailable data, to determine whether exer-cise taken in a green space is more benefi-cial than similar exercise taken in a citygym (20): there are undoubtedly healtheffects of exercise that do not depend ongreen space (21)].The second uncertainty about the psycho-

logical explanation is the absence of evidencethat the measurable rapid short-term psy-chological and physiological changes thatfollow exposure to natural environments(whether specific for such environments ornot) translate into long-term health benefits.In other words, are these short-term psycho-logical effects related in any way to thesuggested health benefits of living close togreen space for prolonged periods (reducedmortality, cardiovascular disease, chronicinflammatory disorders, and depression)(1, 2, 22, 23), or are they a separate tran-sient phenomenon?Summarizing the previous paragraphs,

there is suggestive evidence that living closeto the natural environment (defined here asnonbuilt, including gardens and agriculturalland) has long-term health benefits (1, 2, 8).

Author contributions: G.A.R. wrote the paper.

The author declares no conflict of interest.

This article is a PNAS Direct Submission.

1E-mail: g.rook@ucl.ac.uk.

18360–18367 | PNAS | November 12, 2013 | vol. 110 | no. 46 www.pnas.org/cgi/doi/10.1073/pnas.1313731110

Dow

nloa

ded

by g

uest

on

May

30,

202

0

These benefits might be an additive conse-quence of several effects—an evolved psycho-logical need, plus perhaps exercise, sunlight,and social interactions—but there are noconclusive data. This issue is crucially impor-tant because urban planners need to knowwhether urban green space is really the bestway to achieve the beneficial effects, and as-suming that it is, they then need to know themechanism so that the health advantage de-rived from green spaces can be optimized.We suggest here that humans do indeed havean evolutionarily predetermined need for ex-posure to the natural environment, but thatthis has two distinct components. There is animmunological component that runs in par-allel with the psychological one discussedabove. This discussion is needed becausethe field is split into two distinct trains ofthought that involve different aspects of ourphysiology (the brain and the immune sys-tem) and different scientific and medical dis-ciplines. We hope here to break down theseinterdisciplinary barriers and show that thesetwo pathways are likely to work together inways that can usefully benefit urban planningfor human well-being.

Hygiene Hypothesis and the “OldFriends” MechanismThe high-income countries are undergoingmassive increases in chronic inflammatorydisorders (24–27). The cause is at least partlya failure of immunoregulation, so that theimmune system is attacking inappropriatetargets, such as self, harmless airborne anti-gens (allergens) and gut contents. At birththe immune system is like a computer (an-atomical structures) that contains programs(genetics) but almost no data in terms ofknowledge of molecular structures in theenvironment into which the child is born. Ithas some knowledge of self, acquired aslymphocytes mature in the thymus, andminimal knowledge of the outside world,transferred from the mother across theplacenta. After birth, it needs microbialexposures to provide teaching inputs forseveral crucial reasons. First, exposure toa broad biodiversity of organisms builds upmemory of diverse molecular structuresthat accelerates subsequent rapid recogni-tion of novel dangerous organisms (28, 29).Second, microbial components taken upsystemically from the gut maintain an es-sential background level of activation of theinnate immune system (30). Third, andmost important in the current context, thesystem needs to develop a network of reg-ulatory pathways and regulatory T cells(Tregs) that stop inappropriate immuneattacks on (i) self; (ii) harmless allergens;and (iii) gut contents (Fig. 1). If immuno-regulation fails to stop immune attack on

any of these categories of forbidden targets,the consequences are (i) autoimmune dis-eases such as multiple sclerosis; (ii) allergicdisorders such as hay fever and atopicasthma; and (iii) inflammatory bowel dis-eases such as ulcerative colitis and Crohn’sdisease (Fig. 1).Finally, the immunoregulatory systems

must also turn off inflammatory responsescompletely when they are not needed. Afailure to do this is regularly seen in high-income countries where persistently raisedlevels of C-reactive protein (CRP) are com-mon (discussed in ref. 31). Persistently raisedinflammatory mediators lead to increasedrisk of cardiovascular disease (32) and de-pression (27, 33, 34). In contrast, a longitu-dinal study of adults in a rural low-incomecountry where there is exposure to high mi-crobial burdens in childhood showed thatthey were able to shut off the inflammatoryresponse when there was no need for it (31).More work is needed to discover whether thesame is true in other low-income settings,whether biomass or biodiversity was moreimportant, and whether the effect was at-tributable to bacteria, fungi, protozoa, hel-minths, or other organisms. Fig. 2 illustratessome of the categories of organisms (OldFriends) implicated in driving the immuno-regulatory mechanisms (reviewed and refer-enced in ref. 35). The crucial point is that allthese organisms needed to be tolerated. Somewere part of our physiology (human micro-biota). Others were harmless but inevitablycontaminating food and water (environ-mental microbiota). Similarly, there were

carrier states due to organisms picked upsoon after birth and helminths that persistfor life. Helminthic parasites needed tobe tolerated because, although not alwaysharmless, once they were established inthe host, any effort by the immune systemto eliminate them was futile and merelycaused tissue damage such as elephanti-asis (36). Thus, helminths are powerfullyimmunoregulatory and act as Treg adju-vants. For example, when patients sufferingfrom early relapsing multiple sclerosis (MS)become infected with helminths, the diseasestops progressing, and circulating myelin-recognizing Tregs appear in the peripheralblood (37, 38), an exciting observationthat has led to formal clinical trials (39).This view is now well supported by experi-mental data and molecular mechanisms. OldFriends can be shown to drive immunoreg-ulation and to block or treat models of aller-gies, autoimmune disease, and inflammatorybowel disease (40–42). Some Old Friends(including members of the human gutmicrobiota such as Bacteroides fragilis) ormolecules that they secrete are known tospecifically expand Treg populations (42–45)or to cause dendritic cells (DCs) to switch toregulatory DCs that preferentially drive im-munoregulation (46). However, it is impor-tant to remember that the gut is not the onlysite where immunoregulation can be inducedby macro- and microorganisms. Helminthssuch as blood nematodes that never enter thegut are powerfully immunoregulatory (36),and recent data implicate the skin and air-ways also (29, 47).

Fig. 1. The immune system does not develop normally in the absence of microbial inputs. In addition to a repertoireof potential effector cells, the system also requires regulatory circuits that inhibit damaging responses to inappropriatetargets (such as self, trivial antigens in air, and gut contents) and that terminate inflammatory responses that are nolonger needed. The disease groups that occur when immunoregulation fails are indicated in parentheses.

Rook PNAS | November 12, 2013 | vol. 110 | no. 46 | 18361

PERS

PECT

IVE

Dow

nloa

ded

by g

uest

on

May

30,

202

0

In summary, it is now understood that thevarious classes of organism (Fig. 2) that hadto be tolerated were entrusted by evolution-ary processes with the role of setting up theimmunoregulatory mechanisms and Tregpopulations. If this process fails, we developsusceptibility to chronic inflammatory dis-eases (24, 26, 37, 48), cardiovascular disease(49, 50), and some forms of inflammation-associated depression (27, 33, 51). This con-cept constitutes the Darwinian reformula-tion of the hygiene hypothesis in which theemphasis is on lifestyle changes that reduceour exposure to these immunoregulation-inducing Old Friends.

Progressive Loss of Microbial InputsFig. 2 points out that in modern high-incomesocieties, we have lost many of these catego-ries of immunoregulatory Old Friends, so weare now much more dependent on the im-munoregulation-inducing effects of the mi-crobiota of other humans and on organismsfrom the natural environment. These sourcesof biodiversity are all that remain. However,increasing trends toward agricultural mono-culture are likely to decrease rural microbialbiodiversity because each crop is associat-ed with strikingly different populations of

bacteria, archaea, fungi, protozoa, nemat-odes, etc. (52). Similarly, the chronic in-flammatory disorders that have risenstrikingly in prevalence in developed high-income countries are usually found to bestill more common in urban environmentsfrom which the immunoregulatory OldFriends are essentially absent (Fig. 2). Thisurban increase is true for allergies (53, 54),inflammatory bowel disease (55), and forautoimmune diseases such as MS (56–58).These urban-rural differences are equallyobvious in psychiatric disorders (59). Forexample, a meta-analysis of high-qualitystudies performed in high-income countriessince 1985 found that the prevalence of de-pression in urban areas was 39% higher thanin rural areas. Similarly, the prevalence ofanxiety disorders was 21% higher in urbanthan in rural areas (60), although a smallminority of studies fails to find this urban-rural difference (61). Peen and colleaguesalso noted an increased urban prevalence ofpsychiatric disorders in general (38% morein urban communities) (60), and this isstrikingly true for schizophrenia (62) andautism (63)—both of which involve an in-flammatory component (64). It is usuallysuggested that these results are explained

by greater stressfulness of urban life, orpoor urban social networks, but data toestablish these explanations are not pro-vided (59). Meanwhile, an increasinglystrong case can be made for the involvementof inflammation, secondary to failing im-munoregulation in urban environments (27).It should, however, be noted that not all

studies find higher disease prevalences inurban rather than in rural environments. Astudy performed in the United States foundno link between city-level greenness andheart disease and found greater overall mor-tality in greener cities, possibly attributable togreater car use (17). However, this city-levelstudy did not provide data at the level of theproximity of the individual to green space.Moreover, it combined data from cities incool wet environments with data from deserts,and apart from heart disease, did not docu-ment the chronic inflammatory disorders thatare influenced by the immunological mecha-nisms discussed here.Another situation that leads to a loss of

exposure to microbial biodiversity is immi-gration from a developing country to a high-income urban center. This migration leads torapid loss of the first three categories of or-ganism shown in Fig. 2. In such immigrantpopulations, there are large increases in au-toimmunity (27, 65–69), inflammatory boweldisease (55, 70, 71), depression (27, 72, 73),and allergic disorders (74–77). For allergicdisorders, this has been rigorously docu-mented for children adopted into Swedenfrom low/middle-income countries (75), forMexican immigrants to the United States(76), and for immigrants to Israel from theformer Soviet Union or Ethiopia (77).

Natural Environment andImmunoregulationSome aspects of the Old Friends mechanism(such as immunoregulatory effects of hel-minths) are clearly not directly relevant to thegreen space phenomenon in high-incomecountries. Their absence merely increases ourdependence on immunoregulation-inducingexposures from elsewhere, and the focus ofthis paper is the natural environment.Observations linking exposure to greenenvironments and agriculture to protectionfrom illness were made as early as the 19thcentury by Blackley who noticed that hayfever was rare among farmers (78). This ob-servation has now been confirmed in manycountries using rigorous epidemiologi-cal methods (54, 79). Moreover exposure tofarms also protects from juvenile forms ofinflammatory bowel disease (80). In agree-ment with the theme of this paper, recentstudies indicate that the mechanism of thisprotection from allergic disorders involvesexposure to microbial biodiversity (81).

Fig. 2. A simple classification of organisms and parasites with which humans coevolved and that have been im-plicated by epidemiology or experimental models in the modulation of immunoregulation (although those listed ascarrier states might be biomarkers of exchange of microbiota with other humans). The modern urban environmenteliminates the first three categories, thus increasing our dependence on microbiota from other humans and from thenatural environment.

18362 | www.pnas.org/cgi/doi/10.1073/pnas.1313731110 Rook

Dow

nloa

ded

by g

uest

on

May

30,

202

0

Mattress dust was screened for bacterialDNA (48) and in a separate study samplesof settled dust from children’s rooms weregathered with electrostatic dust samplersand evaluated for bacterial and fungal taxausing culture techniques. In both studies,the diversity of microorganisms found wasinversely related to the risk of asthma (48).Some of the microorganisms found in dustcollected from protective farms have beenshown to exert potent antiallergic effects inanimal models (82–84) and constitute en-vironmental Old Friends. Recently Hanskiand colleagues recorded the skin micro-biota as well as allergic sensitization tocommon allergens in an ecologically mixedarea of Finland. Subjects living close toagricultural land rather than urban agg-lomerations had higher generic diversityof proteobacteria in their skin microbiotaand less atopic sensitization (47). Similarly,a genetically homogeneous population liv-ing in Karelia is partitioned between Fin-land and Russia. In Finnish Karelia, theprevalence of type 1 diabetes is sixfold higherand childhood atopy is fourfold higher thanin Russian Karelia. These differences areassociated with strikingly different microbialpopulations in the home, with much greaterdiversity and many more animal-associatedstrains in Russian homes and more plant-associated species in the Finnish homes (85,86). We cannot at this stage know whetherthe crucial factor is biodiversity, total micro-bial biomass, animal-derived organisms, ororganisms from other environmental sources.

Microbial Diversity and the AirWhat are the microbial exposures that resultfrom proximity to the natural environment orfarms or that fall onto settle plates in a child’sbedroom? First, the air itself contains largenumbers of microorganisms, some of whichmay actively metabolize and replicate in theair (87). Particulate matter in the air such aspollen carries a load of bacteria (88). Manyairborne particles aremore than 5 μmandwilltherefore be deposited in the upper airways, sothat after being carried up the trachea by theaction of cilia, they will be swallowed. There-fore, airborne microorganisms end up on theskin, in the airways, and in the gut where theymodulate the immune system.When total numbers of organisms in air

were counted (i.e., not only the cultivableones) levels of 105/m3 or more were regularlyencountered over a grassy field on clear sunnydays, and estimates approaching 106/m3 havebeen reported above shrubs and some grass-lands (reviewed in ref. 89). The air in facilitieshousing agricultural animals can contain stillhigher numbers, reaching 107–108 archaeaand bacteria/m3 (90). Aerosols collected inTexas contained at least 1,800 different bac-terial types, representing diversity comparable

to that seen in some soils (91). Indeed bacteriacommonly found in soil and water are abun-dant in outdoor air (92, 93). Recent samplesfrom the upper troposphere contained vari-able proportions of bacteria thought to origi-nate from soil, feces, fresh water, or the sea(94). Thus, blue space is another source ofmicrobial biodiversity. Living by the coastdoes yield health benefits (8), and marinespray is a rich source of usually harmlessmicroorganisms (95, 96). Interestingly,an organism that is not harmless providesproof of physiologically significant levels ofintake of marine aerosols and their con-tained life forms. There is rapid (<1 h) onsetof symptoms of brevetoxin poisoning whilewalking on beaches during algal blooms(97) and parallel increases in pulmonaryproblems at some distance inland from suchbeaches (98).The microbial diversity that we encounter

in the natural environment comes mostlyfrom the soil and from plants and from anyanimals that are present (89, 92, 93). Themicrobiota of the soil has huge complexity,and is only now beginning to be explored ina global effort (www.earthmicrobiome.org/)(99). Our ignorance of what is out thereremains profound, and these gaps in ourunderstanding have been referred to as mi-crobial “dark matter” (100). Tens of thou-sands of microbial species are associated withthe rhizosphere (the below ground microbialhabitat constituted by plant root systems)and the phyllosphere (above ground micro-bial habitats provided by plants). The crucialpoint is that plants are able to shape themicrobiota of their rhizospheres (101). Thus,the nature of the vegetation in a green spacewill directly modulate the microbiota presentin the soil, rhizosphere, and phyllosphere(102) and indirectly modulate the microbiotaavailable from coexisting animal life. Thenature, quantity, and diversity of microor-ganisms present is strikingly affected by ag-ricultural practices (52), and it is likely thatthe modern trend toward vast areas of mono-culture will reduce that diversity and furtherdecrease exposure to immunoregulation-inducing organisms in wealthy countries.

Microbiota from Animals and OtherPeopleThe issue of microbiota derived from animalsdeserves further comment. The microbiota indust from households with dogs is signifi-cantly richer and more diverse than thatfound in homes without pets (93, 103). Thisobservation is interesting because exposure topets, particularly dogs, in early life, protectsagainst allergic sensitization and allergic dis-orders (104, 105). Moreover dogs were do-mesticated between 33,000 and 11,000 y ago,so humans have coevolved with dog microbiota

for many millennia (106). Considered togetherwith the protective effects of early exposure tocowsheds (79), other animal-derived strains(85), or animal feces (31), this might suggestthat animals are a particularly importantcomponent of the natural environment.However, other humans are also relevant inthis context. Some consider that social in-teractions are an important consequence ofaccess to the natural environment, and it maybe true that such interactions promote well-being by boosting social capital (13, 14), butthey will also increase the diversity of or-ganisms to which the individual is exposed.Teammates playing a contact sport tended toshare a microbiota, but this converged withthat of the opposing team after a match(107). Similarly cohabiting individuals tendto share microbiota (108). Interestingly peo-ple tend to share even more microbiota withtheir dogs (108), increasing exposure to thedog-associated microbial diversity men-tioned above (93, 103). In sharp contrast,elderly people shut up in care homes withlittle variety in human contact (and littleexposure to the natural environment) havediminished gut microbiota diversity thatcorrelates with poor health outcomes andincreased levels of biomarkers of inflam-mation such as IL-6 (109).

Microbiota of the Built EnvironmentThe airborne microbiota of buildings is stillpoorly defined but can be different from thatof the natural environment. The phylogeneticdiversity of the airborne bacteria in me-chanically ventilated rooms was lower thanthat seen in rooms ventilated via open win-dows or in the outdoor air itself (92, 93).Similarly the organisms derived from soil andwater were abundant in samples of outdoorair but rare or absent from the indoor sam-ples, which were dominated by organismsrelated to human pathogens and commensals(92). Air in cities tends to contain moreparticulate matter, such as diesel particulatesand metallic fragments from subway wheels.Micrococcus species often dominate in urbanair, perhaps associated with such particles(87, 110). The crucial point in the currentcontext is that people living close to agricul-tural land have more biodiverse microbialpopulations on their skin than those livingclose to urban centers (47), and this corre-lates with immunoregulatory differences andreduced atopy. These organisms are also en-countered via the airways and gut. By con-trast, the use of biocides in the home maydecrease microbial biodiversity (111).

Detrimental Microbiota in UnhealthyBuildingsThis point leads on to a further problemwiththe environment in modern buildings. Hu-mans evolved in a natural environment and

Rook PNAS | November 12, 2013 | vol. 110 | no. 46 | 18363

PERS

PECT

IVE

Dow

nloa

ded

by g

uest

on

May

30,

202

0

in contact with animals. Until recently evenour homes were constructed with timber,mud, animal hair, animal dung, thatch, andother natural products and were ventilatedby outside air. By contrast, modern buildingsare constructed with synthetic materials,plastics, and concrete, and the timber andcardboard are treated with adhesives andbiocides, and the buildings are ventilatedby air conditioning systems. When thesemodern structures degrade, become damp,or accumulate condensation in cavity walls,they do not become colonized with the bac-terial strains with which we coevolved. Theybecome habitats for unusual strains that wedid not encounter during our evolutionaryhistory, some of which synthesize toxic mol-ecules that we are unable to inactivate (112,113). Some examples of “sick building syn-drome” have been tentatively attributed toprolonged exposure to these inappropriateairborne microbiota (112, 113).

Environment and the Human MicrobiotaTo what extent does exposure to these variousenvironmental sources of microbial diversitydirectly modulate the human microbiota? Doorganisms from green space become mem-bers of the human microbiota, or are theseorganisms “pseudocommensals” that impingeon the skin, airways, and gut and have in-dependent immunoregulatory properties?Both mechanisms probably occur, althoughthere are rather limited data on these issues.The environment does play an importantrole in the formation and maintenance ofour microbiota. Fig. 3 illustrates severalpotential mechanisms.From birth, our microbiota are constituted

by colonization with organisms from ourmothers, from other social contacts (107,108), and from the environment, and thenfurther modified by factors such as diet andantibiotics (114, 115). Thus, the lifestyle ofthe individual has major effects on thatindividual’s microbiota. The gut microbiotaof children from traditional villages in Bur-kina Faso is totally different from that ofEuropeans (114). An interesting animal ex-periment compared piglets that were housedin a natural outdoor environment with ge-netically similar piglets that had been rearedin a very clean indoor facility. Firmicutes, inparticular Lactobacillus strains, were domi-nant in the gut microbiotas of the outdoorpiglets, whereas the hygienic indoor pigletshad reduced Lactobacillus and more poten-tially pathogenic phylotypes (116). The in-door piglets also had dramatically differentpatterns of gene expression in the ileum,discussed in the next section (116). Werethese effects due to direct colonizationby immunoregulation-inducing organismsfrom the outdoor environment (pathway Ain Fig. 3), or did these organisms fail to

colonize but exerted indirect effects on theimmune system? The answer is unclear, butindirect effects certainly can occur in severalways. Some organisms compete with orantagonize established organisms (pathwayB) and therefore alter the microbiota (117).Others alter the immune system directly(pathway D) or modulate the immune sys-tem in ways that lead secondarily to a changein the host–microbiota relationship, whichin turn leads to changes in the microbiota(pathway C in Fig. 3). The last mechanismis well established in experimental models.Genetic manipulations of the innate immunesystem that have profound effects on im-mune function (such as gene knockout) of-ten operate indirectly by altering the gutmicrobiota. The phenotypic effects can thenbe transferred to WT mice that have notbeen genetically modified, by transferring thealtered microbiota (118, 119). It is the alteredmicrobiota that is the proximate cause ofthe altered immunoregulation (118, 119). Atleast one environmental saprophyte that willnot colonize (and is dead when used in ex-perimental models and in human clinicaltrials), can be shown to evoke immunoregu-latory effects that suppress allergic responseswhether injected s.c. (120) or given orally(121) and also exerts antidepressant-likeeffects on the CNS (122). Probiotic strainssuch as some lactobacilli can at leasttemporarily colonize the gut and inducechanges in the microbiota via a variety ofmechanisms discussed elsewhere (117).

Microbiota Diversity and Regulation ofInflammationThe piglets discussed in the previous section,which had been reared in clean interiorswithout exposure to the natural environment,had different patterns of gene expression inthe ileum, much of it related to the immune

system. For example, they had increased type1 IFN activity, increased MHC class 1, andup-regulation of many chemokines (116),implying a more inflammatory state in theguts of animals whose microbiota had notbeen modified and diversified by exposureto the natural environment. This correlationbetween reduced gut microbial biodiversityand poor control of inflammation is a com-mon finding. Mice exhibit at least twoenterotypes (bacterial ecosystems in the gutmicrobiota), one of which has low bio-diversity and correlates with biomarkers ofinflammation (123). Gut microbiota of lim-ited diversity is also characteristic of humaninflammation-associated conditions such asobesity and inflammatory bowel disease (124,125). Similarly, diminished microbiota bio-diversity in institutionalized elderly peoplecorrelates with diminished health and raisedlevels of peripheral inflammatory markerssuch as IL-6 (109). Adequate microbial in-puts are required to maintain diversity ofthe gut microbiota, and such diversity playsa role in the regulation of inflammation.

Microbial Biodiversity as an EcosystemServiceInterestingly, provision of microbial bio-diversity is not conventionally listed as an“ecosystem service.” Ecosystem services areecologically mediated functions essential tosustaining healthy human societies. Majorreviews of these services do not contain thewords inflammation or immunity (126, 127),despite the fact that the immunoregulatoryroles of microorganisms have been knownfor decades. It is hoped that this perspectivearticle will help to bridge the chasm be-tween ecology and medicine/immunology,so that when ecologists consider how tomaximize the services obtained from

Fig. 3. Multiple ways in which the microbiota of the natural environment can modulate the immune system. Thismodulation may or may not involve colonization. There can be direct interaction with the immune system (pathway D)or indirect effects secondary to changes in the microbiota (pathways A, B, and C, fully explained in text).

18364 | www.pnas.org/cgi/doi/10.1073/pnas.1313731110 Rook

Dow

nloa

ded

by g

uest

on

May

30,

202

0

ecosystems, the contained microbial bio-diversity is taken into consideration.

Psychology vs. ImmunologyThe major conclusion of this review is thatthe beneficial effects of exposure to naturalenvironments are likely to have two separatebut interacting components. First, there arewell-established rapid psychological effectsthat might be explained by an evolved psy-chological reward from contemplating theideal hunter-gatherer habitat. However, thespecificity of the effect for green space hasnot been proven by comparison with otherrelaxing environments, and the relevance ofsuch rapid transient changes for long-termhealth benefits is unknown.However, there is good evidence that the

long-term benefit of exposure to the naturalenvironment is one component of a broadrange of effects that fall under the umbrellaterms hygiene hypothesis or Old Friendsmechanism or biodiversity hypothesis. Theseterms refer to the evolved need for the im-mune system to receive inputs provided bymicrobial biodiversity, and in particular, byorganisms that need to be tolerated, andtherefore have coevolved roles as inducersof the immunoregulatory pathways (26,27). These immunoregulatory mechanismshelp to stop chronic inflammation and itsassociated chronic inflammatory diseases,cardiovascular problems, and depression(27). Unlike the rapid psychological effects,this requires prolonged exposures, particu-larly important during childhood whenmuch of the education of the immune sys-tem occurs. It might not be sufficient toencounter only the biased microbiota of themodern synthetic indoor environment thatlacks the Old Friends and probably bearslittle resemblance to the microbiota we en-countered throughout our evolutionary his-tory. As illustrated in Fig. 2, modern lifedeprives us of many of the inputs that ourimmune systems evolved to anticipate, so weare now more dependent on the microbiotaof other people and the microbiota of thenatural environment and green spaces. Fig. 4provides a list of probable components of thebeneficial effects of the natural environment,with the parallel psychological and immu-nological explanations. It seems likely thatboth types of explanation are important. Theunderlying principle of the immunologicalexplanation is that for many reasons, expo-sure to green spaces will lead to increasedimmunoregulation, resulting in lower back-ground inflammation, manifested as lowerresting CRP. Improved control of in-flammation results in lower prevalence ofinflammatory disorders, cardiovascular dis-ease, and depression and increased stressresilience (27). It is interesting that sunlightand exercise both contribute to this effect.

Sunlight enhances production of vitamin D(128) and nitric oxide (129), both of whichplay critical roles in immunoregulation. Sim-ilarly exercise increases the activity of Tregs(21, 130). Therefore, multiple physiologicalconsequences of exposure to the natural en-vironment will supplement the immunoreg-ulatory effects of microbial biodiversity.

Urgent Questions and Research IssuesMuch research will be needed to consolidatethis immuno-microbiological view of thehealth benefits of exposure to the naturalenvironment. We need new epidemiologicalstudies that concentrate on inflammatorydisorders attributable to defective immuno-regulation. More studies that use high sensi-tivity CRP levels as a surrogate biomarker forbackground inflammation will be valuable.We need more information about the or-ganisms people encounter in their ownhomes and how this is affected by the prox-imity to natural environments and by thenature of that environment. Humans evolvedas a grassland species, so we can guess, forexample, that deserts and some types ofmonoculture might be less beneficial thangrassland and its associated animals. Theexposures of individuals can be monitoredby sampling skin and gut microbiota andperhaps by studying antibodies. We haveworryingly little knowledge about the rela-tionship between environmental strains andthose that colonize humans, because currentmethods of studying microbiota usually giveonly a broad taxonomic grouping.

If it turns out that the immuno-microbi-ological view is correct, we will also need toknow when the educational and immuno-regulatory inputs to our immune systemsneed to occur. At least some of the immu-noregulatory effects of Old Friends are exer-ted very early in life. For example, factors thatdelay (caesarian births) or distort (perinatalantibiotic use) the establishment of the gutmicrobiota of the infant increase the fre-quency of allergic disorders (131, 132). Sim-ilarly, the reduced prevalence of allergicdisorders after exposure to the farm envi-ronment only occurs if the exposure is duringpregnancy or the neonatal period (79, 133).Does this mean that all these immunoregu-latory effects occur during the perinatalperiod? Might it be sufficient to design pre-school daycare centers so that infants areexposed to relevant microbiota? This view isunlikely because later childhood might alsobe important. For example, environmentalfactors, probably microbial, that influence therisk of developing MS in later life seem tooperate during childhood up to the age of10–15 y (65, 68, 69, 134). Therefore, couldthe problem be solved by ensuring the pres-ence of appropriate organisms in homes andschools until after adolescence? This possi-bility also seems unlikely because we knowthat helminth infections, even when they didnot enter the gut, were powerfully immuno-regulatory even in adults (36–38), and there ismounting evidence that dysbiosis or dimin-ished biodiversity of the gut microbiota isassociated with a variety of inflammatory

Fig. 4. Immunological and psychological explanations for the health benefits derived from contact with the naturalenvironment. (NO, nitric oxide). There are many studies of exposures during the perinatal period that point to theimmunological mechanisms, whereas most studies in adult life have been orientated toward psychological explan-ations, and have not included investigation of the immunoregulatory aspects.

Rook PNAS | November 12, 2013 | vol. 110 | no. 46 | 18365

PERS

PECT

IVE

Dow

nloa

ded

by g

uest

on

May

30,

202

0

conditions (109, 124, 125). There is a worry-ing lack of data on this point, but overall, itseems probable that most education of theimmune system occurs in the perinatal periodbut that there is also an ongoing requirementfor microbial inputs throughout life.

ConclusionsIt is interesting that the beneficial effects ofproximity to the natural environment areparticularly prominent in individuals of lowsocioeconomic status (1–3, 8). Perhapswealthier individuals are better able tosupplement such exposures with holidaytravel and rural second homes. This dis-turbing health gradient emphasizes the need

for more research that will enable us todesign urban green spaces that provide notonly the psychological input to our brainsbut also an optimized microbial input toour immune systems. The research out-lined above will help us to know what isneeded and when. If a significant part ofthe role of the natural environment is toprovide an appropriate airborne micro-biota, then multiple, small, widely distrib-uted urban green spaces of high microbialquality might suffice as supplements toa core of large recreational parks. There isalready huge interest in the construction ofroof gardens, vertical gardens and urban

green spaces motivated by aesthetics andby organizations wishing to promote urbanhabitats for birds and insects and by urbanplanners wishing to delay the entry of raindownpours in sewer systems. However, wesuggest that combating the epidemic ofinflammation-associated illnesses in high-income urban environments provides an-other compelling motive for creating greenspaces, and we hope that this paper willenhance collaboration between the medicalprofession, ecologists, and urban planners.

ACKNOWLEDGMENTS. G.A.R. is supported by theNational Institute for Health Research University CollegeLondon Hospitals Biomedical Research Centre.

1 Maas J, Verheij RA, Groenewegen PP, de Vries S, Spreeuwenberg P

(2006) Green space, urbanity, and health: how strong is the relation?

J Epidemiol Community Health 60(7):587–592.2 Mitchell R, Popham F (2008) Effect of exposure to natural

environment on health inequalities: An observational population

study. Lancet 372(9650):1655–1660.3 Dadvand P, et al. (2012) Green space, health inequality and

pregnancy. Environ Int 40:110–115.4 Berman MG, Jonides J, Kaplan S (2008) The cognitive benefits of

interacting with nature. Psychol Sci 19(12):1207–1212.5 Aspinall P, Mavros P, Coyne R, Roe J (2013) The urban brain:

Analysing outdoor physical activity with mobile EEG [published online

ahead of print March 6, 2013]. Br J Sports Med, bjsports-2012-

091877.6 Tsunetsugu Y, Park BJ, Miyazaki Y (2010) Trends in research related

to “Shinrin-yoku” (taking in the forest atmosphere or forest bathing)

in Japan. Environ Health Prev Med 15(1):27–37.7 Park BJ, Tsunetsugu Y, Kasetani T, Kagawa T, Miyazaki Y (2010) The

physiological effects of Shinrin-yoku (taking in the forest atmosphere

or forest bathing): Evidence from field experiments in 24 forests

across Japan. Environ Health Prev Med 15(1):18–26.8 Wheeler BW, White M, Stahl-Timmins W, Depledge MH (2012)

Does living by the coast improve health and wellbeing? Health Place

18(5):1198–1201.9 Ulrich RS (1984) View through a window may influence recovery

from surgery. Science 224(4647):420–421.10 Frumkin H (2001) Beyond toxicity: Human health and the natural

environment. Am J Prev Med 20(3):234–240.11 Morris DW (2011) Adaptation and habitat selection in the eco-

evolutionary process. Proc Biol Sci 278(1717):2401–2411.12 Sponheimer M, et al. (2013) Isotopic evidence of early hominin

diets. Proc Natl Acad Sci USA 110(26):10513–10518.13 Kuo FE, Sullivan WC, Coley RL, Brunson L (1998) Fertile ground

for community: Inner city neighbourhood common spaces. Am J

Community Psychol 26(6):823–851.14 Nyqvist F, Nygård M, Steenbeek W (2013) Social capital and self-

rated health amongst older people in Western Finland and Northern

Sweden: A multi-level analysis. Int J Behav Med, 10.1007/s12529-

013-9307-0.15 Maas J, Verheij RA, Spreeuwenberg P, Groenewegen PP (2008)

Physical activity as a possible mechanism behind the relationship

between green space and health: A multilevel analysis. BMC Public

Health 8:206.16 Oakes JM, Forsyth A, Schmitz KH (2007) The effects of

neighborhood density and street connectivity on walking behavior:

The Twin Cities walking study. Epidemiol Perspect Innov 4:16.17 Richardson EA, et al. (2012) Green cities and health: a question

of scale? J Epidemiol Community Health 66(2):160–165.18 Rosenthal NE, et al. (1984) Seasonal affective disorder. A

description of the syndrome and preliminary findings with light

therapy. Arch Gen Psychiatry 41(1):72–80.19 Hobday RA (1997) Sunlight therapy and solar architecture. Med

Hist 41(4):455–472.20 Thompson Coon JT, et al. (2011) Does participating in physical

activity in outdoor natural environments have a greater effect on

physical and mental wellbeing than physical activity indoors? A

systematic review. Environ Sci Technol 45(5):1761–1772.21 Gleeson M, et al. (2011) The anti-inflammatory effects of

exercise: Mechanisms and implications for the prevention and

treatment of disease. Nat Rev Immunol 11(9):607–615.

22 de Vries S, Verheij RA, Groenewegen PP, Spreeuwenberg P

(2003) Natural environments–healthy environments? An exploratory

analysis of the relationship between greenspace and health. Environ

Plan A 35(10):1717–1731.23 Maas J, et al. (2009) Morbidity is related to a green living

environment. J Epidemiol Community Health 63(12):967–973.24 Bach JF (2002) The effect of infections on susceptibility to

autoimmune and allergic diseases. N Engl J Med 347(12):911–920.25 von Hertzen L, Hanski I, Haahtela T (2011) Natural immunity.

Biodiversity loss and inflammatory diseases are two global

megatrends that might be related. EMBO Rep 12(11):1089–1093.26 Rook GA (2012) Hygiene hypothesis and autoimmune diseases.

Clin Rev Allergy Immunol 42(1):5–15.27 Rook GAW, Lowry CA, Raison CL (2013) Microbial old friends,

immunoregulation and stress resilience. Evol Med Public Health 1(1):

46–64.28 Su LF, Kidd BA, Han A, Kotzin JJ, Davis MM (2013) Virus-specific

CD4(+) memory-phenotype T cells are abundant in unexposed

adults. Immunity 38(2):373–383.29 Naik S, et al. (2012) Compartmentalized control of skin immunity

by resident commensals. Science 337(6098):1115–1119.30 Clarke TB, et al. (2010) Recognition of peptidoglycan from the

microbiota by Nod1 enhances systemic innate immunity. Nat Med

16(2):228–231.31 McDade TW, et al. (2012) Analysis of variability of high sensitivity

C-reactive protein in lowland Ecuador reveals no evidence of chronic

low-grade inflammation. Am J Hum Biol 24(5):675–681.32 Rietzschel E, De Buyzere M (2012) High-sensitive C-reactive

protein: Universal prognostic and causative biomarker in heart

disease? Biomarkers Med 6(1):19–34.33 Gimeno D, et al. (2009) Associations of C-reactive protein and

interleukin-6 with cognitive symptoms of depression: 12-year follow-

up of the Whitehall II study. Psychol Med 39(3):413–423.34 Howren MB, Lamkin DM, Suls J (2009) Associations of

depression with C-reactive protein, IL-1, and IL-6: A meta-analysis.

Psychosom Med 71(2):171–186.35 Rook GAW (2010) 99th Dahlem conference on infection,

inflammation and chronic inflammatory disorders: Darwinian

medicine and the ‘hygiene’ or ‘old friends’ hypothesis. Clin Exp

Immunol 160(1):70–79.36 Babu S, Blauvelt CP, Kumaraswami V, Nutman TB (2006)

Regulatory networks induced by live parasites impair both Th1 and

Th2 pathways in patent lymphatic filariasis: implications for parasite

persistence. J Immunol 176(5):3248–3256.37 Correale J, Farez M (2007) Association between parasite infection

and immune responses in multiple sclerosis. Ann Neurol 61(2):

97–108.38 Correale J, Farez MF (2011) The impact of parasite infections on

the course of multiple sclerosis. J Neuroimmunol 233(1-2):6–11.39 Fleming JO, et al. (2011) Probiotic helminth administration in

relapsing-remitting multiple sclerosis: a phase 1 study. Mult Scler

17(6):743–754.40 Round JL, Mazmanian SK (2009) The gut microbiota shapes

intestinal immune responses during health and disease. Nat Rev

Immunol 9(5):313–323.41 Osada Y, Kanazawa T (2010) Parasitic helminths: New weapons

against immunological disorders. J Biomed Biotechnol 2010:743758.42 Karimi K, Inman MD, Bienenstock J, Forsythe P (2009)

Lactobacillus reuteri-induced regulatory T cells protect against an

allergic airway response in mice. Am J Respir Crit Care Med 179(3):

186–193.

43 Round JL, et al. (2011) The Toll-like receptor 2 pathwayestablishes colonization by a commensal of the human microbiota.

Science 332(6032):974–977.44 Grainger JR, et al. (2010) Helminth secretions induce de novo T

cell Foxp3 expression and regulatory function through the TGF-βpathway. J Exp Med 207(11):2331–2341.45 Atarashi K, et al. (2011) Induction of colonic regulatory T cells by

indigenous Clostridium species. Science 331(6015):337–341.46 Smits HH, et al. (2005) Selective probiotic bacteria induce IL-10-

producing regulatory T cells in vitro by modulating dendritic cell

function through dendritic cell-specific intercellular adhesion

molecule 3-grabbing nonintegrin. J Allergy Clin Immunol 115(6):1260–1267.47 Hanski I, et al. (2012) Environmental biodiversity, human

microbiota, and allergy are interrelated. Proc Natl Acad Sci USA

109(21):8334–8339.48 Ege MJ, et al.; GABRIELA Transregio 22 Study Group (2011)

Exposure to environmental microorganisms and childhood asthma.

N Engl J Med 364(8):701–709.49 Kuiper J, et al. (2007) Immunomodulation of the inflammatory

response in atherosclerosis. Curr Opin Lipidol 18(5):521–526.50 McDade TW, Rutherford J, Adair L, Kuzawa CW (2010) Early

origins of inflammation: Microbial exposures in infancy predict lower

levels of C-reactive protein in adulthood. Proc Biol Sci 277(1684):

1129–1137.51 McDade TW, Hoke M, Borja JB, Adair LS, Kuzawa CW (2013) Do

environments in infancy moderate the association between stress

and inflammation in adulthood? Initial evidence from a birth cohort inthe Philippines. Brain Behav Immun 31:23–30.52 Turner TR, et al. (2013) Comparative metatranscriptomics reveals

kingdom level changes in the rhizosphere microbiome of plants

[published online ahead of print July 18, 2013]. ISME J, 10.1038/

ismej.2013.119.53 Nicolaou N, Siddique N, Custovic A (2005) Allergic disease in

urban and rural populations: Increasing prevalence with increasing

urbanization. Allergy 60(11):1357–1360.54 MacNeill SJ, et al.; GABRIELA study group (2013) Asthma and

allergies: Is the farming environment (still) protective in Poland? The

GABRIEL Advanced Studies. Allergy 68(6):771–779.55 Hou JK, El-Serag H, Thirumurthi S (2009) Distribution and

manifestations of inflammatory bowel disease in Asians, Hispanics,

and African Americans: A systematic review. Am J Gastroenterol

104(8):2100–2109.56 Beebe GW, Kurtzke JF, Kurland LT, Auth TL, Nagler B (1967)

Studies on the natural history of multiple sclerosis. 3. Epidemiologic

analysis of the army experience in World War II. Neurology 17(1):

1–17.57 Antonovsky A, et al. (1965) Epidemiologic Study of Multiple

Sclerosis in Israel. I. An Overall Review of Methods and Findings. Arch

Neurol 13:183–193.58 Lowis GW (1990) The social epidemiology of multiple sclerosis.

Sci Total Environ 90:163–190.59 Sundquist K, Frank G, Sundquist J (2004) Urbanisation and

incidence of psychosis and depression: Follow-up study of 4.4 millionwomen and men in Sweden. Br J Psychiatry 184:293–298.60 Peen J, Schoevers RA, Beekman AT, Dekker J (2010) The current

status of urban-rural differences in psychiatric disorders. Acta

Psychiatr Scand 121(2):84–93.61 Kovess-Masfety V, Lecoutour X, Delavelle S (2005) Mood

disorders and urban/rural settings: Comparisons between two French

regions. Soc Psychiatry Psychiatr Epidemiol 40(8):613–618.

18366 | www.pnas.org/cgi/doi/10.1073/pnas.1313731110 Rook

Dow

nloa

ded

by g

uest

on

May

30,

202

0

62 McGrath J, et al. (2004) A systematic review of the incidence ofschizophrenia: The distribution of rates and the influence of sex,urbanicity, migrant status and methodology. BMC Med 2:13.63 Lauritsen MB, Pedersen CB, Mortensen PB (2005) Effects offamilial risk factors and place of birth on the risk of autism: Anationwide register-based study. J Child Psychol Psychiatry 46(9):963–971.64 Meyer U, Feldon J, Dammann O (2011) Schizophrenia andautism: Both shared and disorder-specific pathogenesis via perinatalinflammation? Pediatr Res 69(5 Pt 2):26R–33R.65 Ahlgren C, Oden A, Lycke J (2012) A nationwide survey of theprevalence of multiple sclerosis in immigrant populations of Sweden.Mult Scler 18(8):1099–1107.66 Cabre P (2009) Environmental changes and epidemiology ofmultiple sclerosis in the French West Indies. J Neurol Sci 286(1-2):58–61.67 Söderström U, Aman J, Hjern A (2012) Being born in Swedenincreases the risk for type 1 diabetes - a study of migration of childrento Sweden as a natural experiment. Acta Paediatr 101(1):73–77.68 Milo R, Kahana E (2010) Multiple sclerosis: Geoepidemiology,genetics and the environment. Autoimmun Rev 9(5):A387–A394.69 Gale CR, Martyn CN (1995) Migrant studies in multiple sclerosis.Prog Neurobiol 47(4-5):425–448.70 Li X, Sundquist J, Hemminki K, Sundquist K (2011) Risk ofinflammatory bowel disease in first- and second-generationimmigrants in Sweden: A nationwide follow-up study. InflammBowel Dis 17(8):1784–1791.71 Carr I, Mayberry JF (1999) The effects of migration on ulcerativecolitis: A three-year prospective study among Europeans and first-and second- generation South Asians in Leicester (1991-1994). Am JGastroenterol 94(10):2918–2922.72 Breslau J, Borges G, Hagar Y, Tancredi D, Gilman S (2009)Immigration to the USA and risk for mood and anxiety disorders:Variation by origin and age at immigration. Psychol Med 39(7):1117–1127.73 Vega WA, Sribney WM, Aguilar-Gaxiola S, Kolody B (2004) 12-month prevalence of DSM-III-R psychiatric disorders among MexicanAmericans: Nativity, social assimilation, and age determinants. J NervMent Dis 192(8):532–541.74 Rottem M, Szyper-Kravitz M, Shoenfeld Y (2005) Atopy andasthma in migrants. Int Arch Allergy Immunol 136(2):198–204.75 Hjern A, Rasmussen F, Hedlin G (1999) Age at adoption, ethnicityand atopic disorder: a study of internationally adopted young men inSweden. Pediatr Allergy Immunol 10(2):101–106.76 Eldeirawi K, et al. (2009) Associations of doctor-diagnosedasthma with immigration status, age at immigration, and length ofresidence in the United States in a sample of Mexican AmericanSchool Children in Chicago. J Asthma 46(8):796–802.77 Pereg D, et al. (2008) Prevalence of asthma in a large group ofIsraeli adolescents: Influence of country of birth and age at migration.Allergy 63(8):1040–1045.78 Blackley CH (1873) Experimental Researches on the Causes andNature of Catarrhus aestivus (Hay-Fever and Hay-Asthma) (BaillièreTindall and Cox, London).79 Riedler J, et al.; ALEX Study Team (2001) Exposure to farming inearly life and development of asthma and allergy: A cross-sectionalsurvey. Lancet 358(9288):1129–1133.80 Radon K, et al.; Chronische Autoimmunerkrankungen undKontakt zu Tieren (Chronic Autoimmune Disease and AnimalContact) Study Group (2007) Contact with farm animals in early lifeand juvenile inflammatory bowel disease: A case-control study.Pediatrics 120(2):354–361.81 Heederik D, von Mutius E (2012) Does diversity of environmentalmicrobial exposure matter for the occurrence of allergy and asthma?J Allergy Clin Immunol 130(1):44–50.82 Debarry J, et al. (2007) Acinetobacter lwoffii and Lactococcuslactis strains isolated from farm cowsheds possess strong allergy-protective properties. J Allergy Clin Immunol 119(6):1514–1521.83 Vogel K, et al. (2008) Animal shed Bacillus licheniformis sporespossess allergy-protective as well as inflammatory properties. J AllergyClin Immunol 122(2):307–312.84 Hagner S, et al. (2013) Farm-derived Gram-positive bacteriumStaphylococcus sciuri W620 prevents asthma phenotype in HDM-and OVA-exposed mice. Allergy 68(3):322–329.

85 Pakarinen J, et al. (2008) Predominance of Gram-positive bacteria

in house dust in the low-allergy risk Russian Karelia. Environ Microbiol

10(12):3317–3325.86 Kondrashova A, et al. (2005) A six-fold gradient in the incidence

of type 1 diabetes at the eastern border of Finland. Ann Med 37(1):

67–72.87 Womack AM, Bohannan BJ, Green JL (2010) Biodiversity and

biogeography of the atmosphere. Philos Trans R Soc Lond B Biol Sci

365(1558):3645–3653.88 Heydenreich B, et al. (2012) Gram-positive bacteria on grass

pollen exhibit adjuvant activity inducing inflammatory T cell

responses. Clin Exp Allergy 42(1):76–84.89 Burrows SM, Elbert W, Lawrence MG, Poeschl U (2009) Bacteria

in the global atmosphere—part 1: Review and synthesis of literature

data for different ecosystems. Atmos Chem Phys 9:9263–9280.90 Nehmé B, et al. (2009) Culture-independent characterization of

archaeal biodiversity in swine confinement building bioaerosols. Appl

Environ Microbiol 75(17):5445–5450.91 Brodie EL, et al. (2007) Urban aerosols harbor diverse and

dynamic bacterial populations. Proc Natl Acad Sci USA 104(1):

299–304.92 Kembel SW, et al. (2012) Architectural design influences the

diversity and structure of the built environment microbiome. ISME J

6(8):1469–1479.93 Dunn RR, Fierer N, Henley JB, Leff JW, Menninger HL (2013)

Home life: Factors structuring the bacterial diversity found within and

between homes. PLoS ONE 8(5):e64133.94 DeLeon-Rodriguez N, et al. (2013) Microbiome of the upper

troposphere: Species composition and prevalence, effects of tropical

storms, and atmospheric implications. Proc Natl Acad Sci USA

110(7):2575–2580.95 Leck C, Bigg K (2005) Biogenic particles in the surface microlayer

and overlaying atmosphere in the central Arctic Ocean during

summer. Tellus 57B(4):305–316.96 Prather KA, et al. (2013) Bringing the ocean into the laboratory

to probe the chemical complexity of sea spray aerosol. Proc Natl

Acad Sci USA 110(19):7550–7555.97 Fleming LE, et al. (2007) Aerosolized red-tide toxins (brevetoxins)

and asthma. Chest 131(1):187–194.98 Kirkpatrick B, et al. (2008) Florida red tide and human health:

Apilot beach conditions reporting system to minimize human

exposure. Sci Total Environ 402(1):1–8.99 Jansson JK, Prosser JI (2013) Microbiology: The life beneath our

feet. Nature 494(7435):40–41.100 Rinke C, et al. (2013) Insights into the phylogeny and coding

potential of microbial dark matter. Nature 499(7459):431–437.101 Berendsen RL, Pieterse CM, Bakker PA (2012) The rhizosphere

microbiome and plant health. Trends Plant Sci 17(8):478–486.102 Kowalchuk GA, Buma DS, de Boer W, Klinkhamer PG, van

Veen JA (2002) Effects of above-ground plant species composition

and diversity on the diversity of soil-borne microorganisms. Antonie

van Leeuwenhoek 81(1-4):509–520.103 Fujimura KE, et al. (2010) Man’s best friend? The effect of pet

ownership on house dust microbial communities. J Allergy Clin

Immunol 126(2):410–413.104 Ownby DR, Johnson CC, Peterson EL (2002) Exposure to

dogs and cats in the first year of life and risk of allergic sensitization at

6 to 7 years of age. JAMA 288(8):963–972.105 Aichbhaumik N, et al. (2008) Prenatal exposure to household

pets influences fetal immunoglobulin E production. Clin Exp Allergy

38(11):1787–1794.106 Axelsson E, et al. (2013) The genomic signature of dog

domestication reveals adaptation to a starch-rich diet. Nature

495(7441):360–364.107 Meadow JF, Bateman AC, Herkert KM, O’Connor TK,

Green JL (2013) Significant changes in the skin microbiome

mediated by the sport of roller derby. PeerJ 1:e53.108 Song SJ, et al. (2013) Cohabiting family members share

microbiota with one another and with their dogs. Elife 2:e00458.109 Claesson MJ, et al. (2012) Gut microbiota composition

correlates with diet and health in the elderly. Nature 488(7410):

178–184.

110 Fang Z, Ouyang Z, Zheng H, Wang X, Hu L (2007) Culturable

airborne bacteria in outdoor environments in Beijing,China. Microb

Ecol 54(3):487–496.111 McBain AJ, et al. (2003) Exposure of sink drain microcosms to

triclosan: Population dynamics and antimicrobial susceptibility. Appl

Environ Microbiol 69(9):5433–5442.112 Andersson MA, et al. (1998) The mitochondrial toxin produced

by Streptomyces griseus strains isolated from an indoor environment

is valinomycin. Appl Environ Microbiol 64(12):4767–4773.113 Sahlberg B, Wieslander G, Norbäck D (2010) Sick building

syndrome in relation to domestic exposure in Sweden—a cohort

study from 1991 to 2001. Scand J Public Health 38(3):232–238.114 De Filippo C, et al. (2010) Impact of diet in shaping gut

microbiota revealed by a comparative study in children from Europe

and rural Africa. Proc Natl Acad Sci USA 107(33):14691–14696.115 Dethlefsen L, Huse S, Sogin ML, Relman DA (2008) The

pervasive effects of an antibiotic on the human gut microbiota, as

revealed by deep 16S rRNA sequencing. PLoS Biol 6(11):e280.116 Mulder IE, et al. (2009) Environmentally-acquired bacteria

influence microbial diversity and natural innate immune responses at

gut surfaces. BMC Biol 7:79.117 Hemarajata P, Versalovic J (2013) Effects of probiotics on gut

microbiota: Mechanisms of intestinal immunomodulation and

neuromodulation. Therap Adv Gastroenterol 6(1):39–51.118 Wen L, et al. (2008) Innate immunity and intestinal microbiota

in the development of Type 1 diabetes. Nature 455(7216):

1109–1113.119 Elinav E, et al. (2011) NLRP6 inflammasome regulates colonic

microbial ecology and risk for colitis. Cell 145(5):745–757.120 Zuany-Amorim C, et al. (2002) Suppression of airway

eosinophilia by killed Mycobacterium vaccae-induced allergen-

specific regulatory T-cells. Nat Med 8(6):625–629.121 Hunt JR, Martinelli R, Adams VC, Rook GAW, Brunet LR (2005)

Intragastric administration of Mycobacterium vaccae inhibits severe

pulmonary allergic inflammation in a mouse model. Clin Exp Allergy

35(5):685–690.122 Lowry CA, et al. (2007) Identification of an immune-responsive

mesolimbocortical serotonergic system: Potential role in regulation of

emotional behavior. Neuroscience 146(2):756–772.123 Hildebrand F, et al. (2013) Inflammation-associated

enterotypes, host genotype, cage and inter-individual effects drive

gut microbiota variation in common laboratory mice. Genome Biol

14(1):R4.124 Turnbaugh PJ, et al. (2009) A core gut microbiome in obese and

lean twins. Nature 457(7228):480–484.125 Rehman A, et al. (2010) Transcriptional activity of the dominant

gut mucosal microbiota in chronic inflammatory bowel disease

patients. J Med Microbiol 59(Pt 9):1114–1122.126 Reid WV (2005) Millennium Ecosystem Assessment, 2005.

Ecosystems and Human Well-being: Synthesis (Island Press,

Washington, DC).127 Mooney H, Cropper A, Reid W (2005) Confronting the human

dilemma. Nature 434(7033):561–562.128 Milaneschi Y, et al. (2013) The association between low vitamin

D and depressive disorders. Mol Psychiatry, 10.1038/mp.2013.36.129 Feelisch M, et al. (2010) Is sunlight good for our heart? Eur

Heart J 31(9):1041–1045.130 Lowder T, Dugger K, Deshane J, Estell K, Schwiebert LM (2010)

Repeated bouts of aerobic exercise enhance regulatory T cell

responses in a murine asthma model. Brain Behav Immun 24(1):

153–159.131 Thavagnanam S, Fleming J, Bromley A, Shields MD,

Cardwell CR (2008) A meta-analysis of the association between

Caesarean section and childhood asthma. Clin Exp Allergy 38(4):

629–633.132 Stensballe LG, Simonsen J, Jensen SM, Bonnelykke K,

Bisgaard H (2013) Use of antibiotics during pregnancy increases the

risk of asthma in early childhood. J Pediatr 162(4):832–838.133 Ege MJ, et al. (2008) Prenatal exposure to a farm environment

modifies atopic sensitization at birth. J Allergy Clin Immunol 122(2):

407–412.134 Cabre P (2007) Migration and multiple sclerosis: The French

West Indies experience. J Neurol Sci 262(1-2):117–121.

Rook PNAS | November 12, 2013 | vol. 110 | no. 46 | 18367

PERS

PECT

IVE

Dow

nloa

ded

by g

uest

on

May

30,

202

0